Local Remote Neural Network Authentication Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine learning-based user authentication systems require constant online connectivity for data capture and analysis, which is not always available, especially in environments like underground car parks where internet access is limited.
Innovation Solution
A method utilizing a local machine learning engine with artificial neural networks to process various input signals for authentication, which includes determining likelihood estimations and transmitting signals to a remote server for additional learning phases, allowing for improved authentication efficiency even without continuous online connectivity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning technology is used for authentication, then authentication accuracy is improved, but online connectivity requirement increases
Solution Approach 1:
The machine learning system is segmented into two parts: a local machine learning engine that operates independently on the communication device, and a remote server that provides additional learning capabilities when available. This segmentation allows the system to maintain authentication functionality without continuous online connectivity while still benefiting from remote learning resources when connected.
Solution Approach 2:
The local machine learning engine is pre-trained with initial authentication models before the device is deployed or before online connectivity is lost. This preliminary training enables the device to perform accurate authentication locally without requiring real-time connection to remote servers, resolving the contradiction between maintaining accuracy and reducing connectivity requirements.
2Reliability
If local machine learning engine is implemented, then authentication can be performed offline, but learning capability is limited compared to remote server
Solution Approach 1:
When the communication device reconnects to the remote server, the local machine learning engine receives feedback in the form of additional training data and updated models from the server. This feedback mechanism allows the local engine to continuously improve its learning capability while maintaining offline operational reliability, bridging the gap between independent operation and advanced learning capabilities.
Solution Approach 2:
The system performs preliminary training of the local machine learning engine using available data before offline operation begins. This preliminary preparation ensures that the local engine has sufficient learning capability to operate reliably offline, while the architecture remains open to future improvements when connectivity is restored.
3Measurement precision
If remote server is used for continuous learning, then model accuracy improves, but data transmission requirements increase
Solution Approach 1:
The system extracts only the essential learning updates and model improvements from the remote server rather than transmitting entire datasets or complete models. By taking out only the necessary updates, the system achieves continuous model accuracy improvement while minimizing data transmission volume, directly addressing the contradiction between accuracy and transmission requirements.
Data Source
AI summary
Method for improving user authentication efficiency performed by a communication device belonging to an authentication system. The communication device includes a local machine learning engine having a set of N artificial neural network ANN1,i adapted to process N different types of input signals. The method includes the steps: receiving a first set of N input signals S_1(i) for authentication purpose; determining respectively for each of the N input signals S_1(i) by the N artificial neural networks ANN1,i, N estimations LH(i) of the likelihood that a given input signal is provided by a legitimate user; determining based on a risk scoring established using the N estimations LH(i) if the requesting user is authenticated as the legitimate user; if the requested user is authenticated, determining if at least one likelihood estimation determined for a given input signal S_1(j) is below a predetermined threshold TC,i, and if it is the case: transmitting the input signal S_1 (j) to a remote server implementing a server machine learning engine adapted to process said N different types of input signals and trained to identify a user U_C; receiving an input signal S_2(j) associated to the closest candidate U_C and executing by the local machine learning engine an additional learning phase using the input signal S_2(j) as an input signal that is not associated to the requesting user.